Swedish National Infrastructure for Computing (SNIC)
The Swedish National Infrastructure for Computing (SNIC) is a national research infrastructure that makes available large-scale high-performance computing resources, storage capacity, and advanced user support, for Swedish research.
The field of scientific computing is rapidly developing where programmable graphical processing units (GPUs) are doing much of the modern calculations associated with machine learning and artificial intelligence and many scientific codes written prior to the GPU era requires to be partly re-written (ported) to benefit from the novel architectures available today. Also, data storage has recently expanded with many initiatives taken towards open data, FAIR data and handling of sensitive personal data to address the flow of research data: from generation to analysis, long-term storage and metadata curation for re-usage prior to final archiving.
The Swedish Research Council decided to terminate SNIC funding from 2023 based on an international review published in 2021, while the SNIC user community shared a more positive view on SNIC services - see SNIC facts below.
Today in autumn 2021 the Swedish Research Council and university representatives are discussing how to merge all e-Science infrastructures into a single authority to have a less fragmented compute, storage and training landscape from 2023 and onward.
SNIC is having impact at Karolinska Institutet with 173 Principal Investigators running 216 projects across 19 of 22 KI departments (data from 2019). SNIC is also the partner towards the Wallenberg foundation that recently launched Data Driven Life Science (DDLS) and "Wallenberg Artificial Intelligence, Autonomous Systems and Software Program" (WASP) to support Swedish research and industry with competence in artificial intelligence, autonomous systems and software. SNIC is providing the infrastructure for WASP and DDLS in shape of SNIC resource Alvis and NSC-Berzelius.
SNIC current events and symposiums
No current events announced at the moment
SNIC GPU resources
- Kebnekaise at SNIC-HPC2N, have GPUs available for computing since many years
- Tetralith at SNIC-NSC,retrofit with one Nvidia T4 GPUs per compute node
- Alvis at SNIC-C3SE is a new GPU resource for AI
- BerzeLiUs at SNIC-NSC is a new GPU resource for industrial AI
- Bianca is a research system dedicated for analyzing sensitive personal data.
SNICs compute resource for sensitive personal data consists of a compute cluster (Bianca) and a storage cluster (Castor) hosted by UPPMAX in Uppsala and the service to apply for is called SNIC-SENS. Researchers new to HPC are recommended to apply for a SNIC small compute project and advance to a medium and large if applicable.
The majority of SNIC compute resources are accessible from any Linux/mac/pc via Thinlinc, a remote desktop solution that enables remote graphics and "my own Linux machine" feeling. For some software, such as MATLAB, you may run a graphical user interface at the login node and submit compute jobs to the compute nodes:
SNIC Science Cloud is a large scale, geographically distributed OpenStack cloud Infrastructure as a Service (IaaS), intended for Swedish academic research. SNIC Science Cloud is available free of charge to researchers at Swedish higher education institutions through open application procedures.
SNIC storage allocations must be applied for separate from the compute allocations.
- Centre Storage i.e. storage of active data at the six Swedish compute centers.
- Swestore, subdivided into Swestore-iRODS and Swestore-dCache
Please complement your SNIC compute project by applying separately for SNIC storage by entering "Storage rounds" at the overview of SNIC services. SNIC storage divides storage into small, medium and large and also links to a few local storage resources that are not allocated via SNIC.
EuroHPC is an EU initiative where 32 European countries and 2 private partners are engaged to develop a World Class Supercomputing Ecosystem in Europe. The EuroHPC compute resources are presented here and online and Karolinska Institutet researchers may apply for guidance and proposal writing assistance through "SNIC dedicated user support" or ENCCS proposal support.
The annual EuroHPC calls are published and administrated by EU and also published at vr.se when cofunded by the Swedish Research Council (SRC). In the EuroHPC governance board (GB) Sweden is represented by staff from Department of Education (David Edvardsson) and SRC (Magnus Friberg). The GB decides on work plans with calls and national funders (like SRC) may participate based on funding situation and scope. EuroHPC calls aim for consortia to apply and a few examples of previously funded life-science projects are:
EuroHPC Joint Undertaking organise evaluations of applications and the governance board (GB) decide, followed by negotiations and grant agreement between EU and the consortia. Successful consortia achive funding from EuroHPC and co-funding from member states. Please contact SRC and Swedish governance board representatives for support and questions.
ENCCS and LUMI pre-exascale supercomputer
Sweden contributes to EuroHPC with a competence center acronymed ENCCS, EuroCC National Competence Center Sweden, that offers e-Science training and adaption of selected software towards (pre)-exascale systems. ENCCS is crucial to make the Swedish research community ready for LUMI the most powerful EuroHPC resource.
Sweden participates in the LUMI consortium via the Swedish Research Council and SNIC will handle Swedish LUMI projects such as calls, reviews, allocation and statistics by connecting SUPR (SNIC User and Project Repository) to LUMI via the Puhuri project. The dedicated part of LUMI available to Swedish researchers corresponds to our funding being 3.5% of the LUMI resource. Please note Swedish researchers might get a larger share of LUMI if successful in the international calls to be managed by PRACE. The established PRACE application system will used for international calls by EuroHPC and ENCCS offers proposal support.
The allocations for the Swedish part of LUMI will divided into
- General access (Tier-1 Project Access)
- Extreme scale (Tier-0 Project Access)
where the first General access and Extreme scale calls will open in the autumn of 2021. The national Extreme scale proposal templates will be identical to the international Extreme scale proposal templates to encourage Swedish researchers applying to the international calls and ease coordination between national and international EuroHPC and LUMI projects.
The quality and number of published research articles is a common way to assess the impact and success of scientific work. SNIC therefore explicitly requests the acknowledgement of SNIC in scientific publications and presentations that required the use of SNIC resources, services or expertise.
SNIC acknowledgement templates
- "The computations/simulations/[SIMILAR] were performed on resources provided by the Swedish National Infrastructure for Computing (SNIC) at [SNIC CENTER]."
If applicable, share an acknowledgement for application support:
- "[NAME] at [SNIC CENTER] is acknowledged for assistance concerning technical and implementational aspects [OR SIMILAR] in making the code run on the [OR SIMILAR] [SNIC CENTER] resources."
- "We thank [NAME] at [SNIC CENTER] for [his/her] assistance with [describe tasks such as porting, optimization, etc.], which was made possible through application support provided by SNIC."